Audit health-check artifacts and runtime state for staleness and quality gaps.
npx skills add https://github.com/vTRKA/supervibe --skill audit-vtrka
Please help me install this Agent Skill. Skill: audit Source: https://github.com/vTRKA/supervibe/tree/main/skills/audit Command: npx skills add https://github.com/vTRKA/supervibe --skill audit-vtrka
Audit prevents stale, unverified, or low-quality project changes by checking that health-check signals and evidence-backed artifacts are present before an agent proceeds.
Ask your AI to audit the project for stale or weak evidence by running /supervibe-audit and stopping once the structured health report is returned.
High-intent search queries and answers about installing and using this skill.
Project audit ensures AI-assisted coding proceeds only with evidence-backed context by checking health-check artifacts and runtime state for staleness. It detects coverage gaps, stale references, and weak artifacts, emitting a bounded PASS, FAIL, PARTIAL, BLOCKED, or ADVISORY decision with confidence caps.
You should trigger a project audit at session start, after substantial file changes, or when an agent reports stale-context blockers. This ensures project memory, indexing, receipts, and validators are checked for staleness and quality gaps before any AI-assisted coding proceeds.
No, the project audit is strictly non-mutating and does not modify production artifacts. It emits a structured health report with cited evidence, confidence caps, blockers, and nextAction recommendations, providing a bounded decision without altering your codebase.
A project health audit reads registry entries, memory index state, confidence and effectiveness logs, receipts, and recent context. It detects stale references, coverage gaps, weak artifacts, stale agent verifications, and excessive override rates to produce a bounded audit decision.
Stale-context blockers occur when health-check artifacts, memory index state, or evidence receipts are outdated after substantial file changes. Running a project audit verifies findings with reproducible evidence, identifies coverage gaps, and clears blockers by emitting a structured health report with confidence caps.
user-no-foundCreate a Record directory with draft-plan.md and Memory entries for new projects.
vTRKAConvert weak or stale project artifacts into evidence-backed updates with dry-run diffs.
krzemienskiRun parallel validators on identical journeys and synthesize a consensus report.
daemon-blockint-techValidate adversary hypotheses across enterprise telemetry using MITRE ATT&CK mapping.
